การหาตำแหน่งรอยแตกร้าวด้วยคลื่นอะคูสติกและระบบเครือข่ายประสาทเทียม

Abstract

This article presents the application of Neural network in continuous structural monitoring with acoustic emission technology. The result indicates that it is only the velocity of the acoustic wave of Hsu-Nielsen source that can be used as data to train the neural network in learning the relationship between the characteristics of acoustic wave and the location of the crack. The experiments with plain steel plate and steel plate with stiffeners have shown that the proposed method is able to correctly identify the location of the crack and is insensitive to noise and error in measurement.